Ex vivo evaluation of the biomechanical effect if varying monocortical screw numbers on a plate-rod canine femoral gap model

Objective: To compare the biomechanical behaviour of plate-rod constructs with varying numbers of monocortical screws applied to an ex vivo canine femoral-gap ostectomy model.

Sample population: Twenty Greyhound dog cadaveric femurs.

Methods: Bone mineral density (BMD) was assessed with dual x-ray absorptiometry. Bones were assigned to four groups. Bones had a 12-hole 3.5 mm locking compression plate with one bicortical non-locking cortical screw in the most proximal and distal plate holes and an intramedullary Steinmann pin applied across a 20 mm mid-diaphyseal ostectomy. Additionally, one to four monocortical non-locking cortical screws were then placed (Groups 1-4 respectively) in the proximal and distal fragments. Stiffness and axial collapse were determined before and after cyclic axial loading (6000 cycles at' 20%, 40%, and 60% of mean bodyweight [total: 18000 cycles]). Constructs subsequently underwent an additional 45000 cycles at 60% of bodyweight (total: 63000 cycles). Loading to failure was then performed and ultimate load and mode of failure recorded.

Results: The BMD did not differ significantly between groups. Construct stiffness for group 1 was significantly less than group 4 (p = 0.008). Stiffness showed a linear increase with an increasing number of monocortical screws (p = 0.001). All constructs survived fatigue loading. Load-to-failure was not significantly different between groups. Mean load- to-failure of all groups was >1350N.

Clinical relevance: Ex vivo canine large-breed femurs showed adequate stability biomechanically and gradually increasing stiffness with increasing monocortical screw numbers.